Comment by jaustin

Elop was the trojan horse that killed Nokia. He worked at MSFT prior to this and single handedly destroyed Nokia. The N9 was revolutionary on its own; GPU accelerated UI, sleek looks, Maemo OS, it is a device people would actually want over the limping Symbian that never fully adapted to touch-only, or the useless Windows Phone 7.

No developer dared to touch Maemo because its future was so uncertain due to the switch to WP. My dad was one of the early Lumia adopters and it was so limiting in what it can do.

I bought an N9 in 2011 and it was an incredible phone. The design and UI were gorgeous and it was a joy to use. I still miss the swipe-driven UI - it was clever, intuitive and well thought out. The phone itself had Facebook, Twitter, WhatsApp, and Spotify clients, and MS Exchange support for calendaring and email (I believe Nokia developed or ported many of these in-house) and was completely usable day-to-day.

Compared to Nokia's symbian phones and earlier Maemo efforts, it felt revolutionary and I'd agree Nokia had a device which could have paved the way for a post-symbian future. It definitely felt like, with continued investment, it would have been a real iPhone competitor, and in just the nick of time.

Elop's strategy was a disaster.

There was even the Qt strategy for making the transition smoother (and better hedged) by having apps portable across the different OSs. It was of course killed too because it could have challenged Windows Phone.

I worked in Nokia at the time and played with the N9. Meego was actually really good. It could have been competitive with the iPhone and Nokia could have stayed at the top and been where Samsung is now.

That was a big source of contention, but admittedly there was plenty of skunkworks going around internally to experiment with the officially forbidden material.

I was probably one of the first people to ever possess[ß] a Nokia device running Linux. A research unit in US wrote a library to interface with the baseband modem and provided the whole thing as a single, mostly-statically linked binary that could be used for phonecall functionality. A skunkworks team in Finland wrote a bootloader for N95 to use a Linux kernel. And an ex team member helped put together the initial Debian-based userland.

I wrote the wrapper library that under the hood ran the baseband binary, exposing a sane state machine you could then rely on from "regular" userspace. And I wrote the first, really rough contact book to make/receive calls from the prototype UI. The UI was built with a very early version of libflutter, a GL-based widget library. We built our own layer on top of it.

The prototype became known as the "Flower Phone", thanks to its default background screen. A few months after the device having been showed off our team was provided with about a dozen bright orange[0] N95 devices that ran Linux, booted off of a userspace we had built, and came with our prototype UI. We used them for on-device debugging and developing the UI layer further. Making real calls with them was a core piece of functionality.

From what I understand, the phone functionality in N900 became a reality thanks to that little project.

ß: wasn't mine, it firmly remained property of Nokia. But I used it for experimentation and making real calls.

0: the colour was used to signal the devices were prototypes.

Oh. That brings so much into perspective. They wouldn't cannibalize their own sales, so someone else did. Classic. How deeply Kodak of them.

If you don’t eat your own lunch, someone else will…

>Nokia had deeply dysfunctional internal politics at that point.

Tell me a large company other than Apple that wasn't completely dysfunctional.